Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Electromotive force wikipedia , lookup
Lorentz force wikipedia , lookup
Maxwell's equations wikipedia , lookup
Skin effect wikipedia , lookup
Earth's magnetic field wikipedia , lookup
Giant magnetoresistance wikipedia , lookup
History of geomagnetism wikipedia , lookup
Electromagnet wikipedia , lookup
Force between magnets wikipedia , lookup
Magnetotellurics wikipedia , lookup
Electromagnetic field wikipedia , lookup
Ferromagnetism wikipedia , lookup
Unit 18 Persistent currents and dynamic effects Helene Felice, Soren Prestemon Lawrence Berkeley National Laboratory (LBNL) Paolo Ferracin and Ezio Todesco European Organization for Nuclear Research (CERN) USPAS June 2009, Superconducting accelerator magnets QUESTIONS Each conductor in a magnetic field tends to shield itself from the field – what happens for superconducting cables in a magnet ? Is there a dependence on ramp rate, and on the history of the magnet ? Are these effect large ? How can we remove them or live with them ? These slides heavily rely on the lecture “Dynamic effects in superconducting magnets” by A. Jain at USPAS, Phoenix 2006, and on chapter 6 of Schmuser USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.2 CONTENTS 1. Persistent currents 2. Decay 3. Snap-back 4. Ramp effects USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.3 1. PERSISTENT CURRENTS: SUPERCONDUCTOR IN EXTERNAL FIELD Superconductor slab in external field A superconductor tries to shield itself from a magnetic field The shielding is made through currents that flow in each filament These currents flow inside the superconductor and therefore have no resistance (persistent) B B - + + + B + + To maximize the shielded area, the persistent currents must have the largest possible current density, i.e. the critical current – which depends on the external field (critical state model) These induced currents perturb the field harmonics USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.4 1. PERSISTENT CURRENTS: SUPERCONDUCTOR IN EXTERNAL FIELD • Superconductor slab in external field For low external fields, the critical current flows in the outer skin to generate a field opposite to the external one – inside the field is zero B B - + + + B + + B When the external field rises, the area filled with shielding currents increases USPAS June 2009, Superconducting accelerator magnets B - + + B + + + Unit 15: Persistent currents and dynamic effects – 15.5 1. PERSISTENT CURRENTS: SUPERCONDUCTOR IN EXTERNAL FIELD Superconductor slab in external field B When the external field reaches a given value (penetration field), all the superconductor is filled with shielding currents B - + + B + + + B Beyond this value, the superconductor cannot do anything more to shield itself and a magnetic field is present inside it USPAS June 2009, Superconducting accelerator magnets B - + + B + + + Unit 15: Persistent currents and dynamic effects – 15.6 1. PERSISTENT CURRENTS: FILAMENT IN EXTERNAL FIELD Circular filament in external field Hypothesis: currents flow in an elliptical shell (Bean model) A bit of trigonometry gives the shielding condition determining b B a - B B 2 0 j c a cos 1 sin cos + B + b b a B At full penetration one has b=0 and the magnetic field is the penetration field B B pf USPAS June 2009, Superconducting accelerator magnets 2 0 jc a a B - + - + B Unit 15: Persistent currents and dynamic effects – 15.7 1. PERSISTENT CURRENTS: FILAMENT IN EXTERNAL FIELD Magnetization in a circular filament in external field Magnetization estimate in the Bean model M 2 0 j c a 2 a2 y2 4 0 j c a b 2 a dy 2 xdx2 3 1 a 2 b 1 y / a a B a - B + B + b At full penetration the magnetization is M M pf 4 0 jc a 3 B a B - + - + B Mr. Bean, English (born on 6 January 1955) USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.8 1. PERSISTENT CURRENTS: PENETRATION FIELD Estimating penetration field Filament size is ~ 5-100 m (see Unit 4) 60000 2 Current density (A/mm ) Critical current surface vs field use parameterization for low j Critical current vs penetration field jc=/(2a) Bp 50000 40000 30000 jc(B) 20000 10000 0 0.0 Filament diameter Temperature Critical current Penetration field Tevatron dipole HERA dipole RHIC dipole LHC dipole (mm) 0.009 0.015 0.007 0.007 (K) 4.2 4.2 4.2 1.9 (A/mm2) 25000 21667 26786 35714 0.2 0.4 0.6 Field (T) 0.8 1.0 (T) 0.09 0.13 0.08 0.10 The penetration field is in general rather low – at high field filaments are fully penetrated USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.9 1. PERSISTENT CURRENTS: RAMPING DOWN The phenomenology in a circular filament in external field ramping down B We reached the condition of full penetration a B If now we ramp down, the external layer of the superconductor will have opposite current to continue the shielding - + - + B B The shielding feature depend not only on the field but also on the previous fields: hysteresis effect USPAS June 2009, Superconducting accelerator magnets a B + - + - + - + - B Unit 15: Persistent currents and dynamic effects – 15.10 1. PERSISTENT CURRENTS: HYSTERESIS Hysteresis of the magnetization From A. Jain, USPAS 2007, Dynamic effects in superconducting magnets, pg. 18 USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.11 1. PERSISTENT CURRENTS: SUMMARY Persistent currents depend on External parameters: magnetic field and its previous values – but not on the rate of change of the field ! Filament parameter: size and geometry Superconductor parameter: critical current Parametric dependence For low external fields, the critical current is larger and therefore the persistent current and induced magnetization are larger At large external fields the critical current becomes smaller and therefore the effect is smaller What happens in a magnet ? Magnetic field in the coil has large variations in module and direction Filament magnetization induces a perturbation in main field and harmonics For larger fields, the harmonics are further reduced by the normalization USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.12 1. PERSISTENT CURRENTS: MEASUREMENTS Measurements of the magnetization Magnetization measurement in HERA cable From P. Schmuser, pg. 85 Fig. 6.3 USPAS June 2009, Superconducting accelerator magnets Magnetization measurement in LHC cable Courtesy of L. Bottura et al. Unit 15: Persistent currents and dynamic effects – 15.13 1. PERSISTENT CURRENT COMPUTATION A way to compute persistent currents Calculate the field map in each strand of the coil (and its previous values) Estimate the critical current vs field using measurements or a parameterization – note that values for low fields are very relevant Compute the persistent currents of the filament using a geometrical model, and scale them to the strand Scale the magnetization to take into account of the transport current – it must flow somewhere ! jt Reduction factor is 1 jc Evaluate the effect of these additional currents to the main field and to the field harmonics Several codes can do this evaluation – Roxie, … USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.14 1. PERSISTENT CURRENT COMPUTATION Comparison measurements vs model A good agreement is found – agreement at 90% Persistent current measured vs computed in HERA dipoles and quadrupoles - From P. Schmuser, pg. 87 Fig. 6.5 Persistent current measured vs computed in Tevatron dipoles From P. Bauer et al, FNAL TD-02-040 (2004) Spread of persistent current given by Differences in the critical current Differences in the filament geometry, deformed after cabling USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.15 1. PERSISTENT CURRENT CORRECTION Even with the finest filaments (5 m) the persistent current give several units in allowed multipoles at injection How to cure this effect ? Several strategies Change the hardware (not cheap …) Of the cable: further reduce the filament size – feasible ? Of the machine: reduce the energy range, i.e. increase the injection current Compensation Design compensation: optimize coil geometry, such that harmonics are minimized at injection Drawback: at high field harmonics will be not optimized, but the beam is smaller … Active compensation through correction magnets (as in HERA, LHC) Passive compensation through ferromagnetic shims or ferromagnetic cold bore USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.16 CONTENTS 1. Persistent currents 2. Decay 3. Snap-back 4. Ramp effects USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.17 2. DECAY Phenomenology The injection of the beam is not instantaneous - takes minutes Tevatron: 30 minutes to 2 hours LHC: 20 minutes HERA: around 20 minutes During the injection the field is constant, and one observes a decay of main field and multipolar components units First observed in Tevatron in 1987 [R. Hanft et al., Appl. Supercond. Conf. (1988), also in TM-1542] Equations proposed to fit the decay Logarithmic decay (HERA, Tevatron) b3 (t ) A R log t Double exponential (RHIC, LHC) b3 (t ) A1 exp( t / t1 ) A2 exp( t / t 2 ) USPAS June 2009, Superconducting accelerator magnets Decay of b3 in HERA dipoles versus time, from Schmuser pg. 90, fig 6.8 Unit 15: Persistent currents and dynamic effects – 15.18 2. DECAY Origins Thermally activated flux creep inducing a decrease in the critical current density It is temperature dependent Produces a logarithmic decay Boundary induced coupling currents The strands carry different currents Current redistribution can affect the magnetization due to the changes in the local field There are indications that both mechanisms are involved in the decay of magnetization USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.19 2. DECAY Dependence on the previous history The amplitude of the decay depends on the parameters of the previous cycles Proportional to the flat-top current of the previous cycle Decreases for longer back-porch Saturates for flat-top duration longer than 1 h The cycle used for powering Tevatron dipoles, from P. Bauer et al, FNAL TD-02-040 (2004) USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.20 CONTENTS 1. Persistent currents 2. Decay 3. Snap-back 4. Ramp effects USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.21 3. SNAPBACK Phenomenology At the end of the injection, the beam is accelerated and the field is ramped up In that moment, the decay of persistent currents disappears and the previous values are recovered Snapback phenomenology in RHIC dipoles, from A. Jain, USPAS 2006, « Dynamic effects and … », slide 27 USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.22 3. SNAPBACK Snapback versus current The snapback versus current (i.e. versus field, and versus time) dependence is exponential I I inj b3sb b3sb exp I The snapback takes place in a few seconds – very fast phenomena Snapback in b3 versus time, LHC main dipoles, from L. Bottura et al, IEEE Trans. Appl. Supercond. 15 (2005) 1217-20. USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.23 3. SNAPBACK Snapback versus current sb 3 b I I inj b exp I sb 3 Two parameters: amplitude (the same of the decay) and “time” constant I b3 I Measurements show that they are proportional The constant depends on the magnet design One can justify this proportionality through a model sb Linear relation between snapback constants for Tevatron (right) and LHC (left) dipoles from L. Bottura et al, IEEE Trans. Appl. Supercond. 15 (2005) 1217-20. Unit 15: Persistent currents and dynamic effects – 15.24 USPAS June 2009, Superconducting accelerator magnets CONTENTS 1. Persistent currents 2. Decay 3. Snap-back 4. Ramp effects USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.25 4. RAMP EFFECTS The ramping of the magnet induces a variation of the flux with time in loops made by strands This variable flux can induce currents which are Proportional to the ramp rate Proportional to the area of the loop Inversely proportional to the inter-strand cross-contact resistance These currents may perturb the field homogeneity for high ramp rates A cure: increase the inter-strand resistance by special coating For example, in the LHC dipoles the ramp rate is 10 A/s, the interstrand resistance is > 15-40 , and the impact on field quality is negligible Another effect can come from flux variation in loops made by filaments Usually called coupling currents USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.26 CONCLUSIONS Persistent currents The mechanism The conductor shields itself from the external magnetic field this generates shielding currents inside the conductor field perturbation Features Contribution is very relevant at injection, disappears at high field Contribution on allowed harmonics, proportional to filament size, gets worse with large energy sweep Contribution depends on the previous values of the magnetic field, but not on the rate of magnetic field change (hysteretic phenomena) Reliable models can predict the persistent currents Corrections Passive correction with ferromagnetic shims Smaller filament, smaller energy sweep Active correction with corrector magnets USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.27 CONCLUSIONS Decay The mechanism When the field is constant (this happens at injection energy) the magnetization decays with time due to different mechanisms Flux creep Boundary induced coupling currents Features Scale time is 100 s Semi-empirical fits with exponential or logarithms are used Contribution depends on the previous history, including the rate of magnetic field change (dynamic effect) No quantitative predictions available Corrections Phenomena are slow, corrector magnets are used to compensate on the fly USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.28 CONCLUSIONS Snapback The mechanism When the field is ramped up again, all the decay of persistent currents is wiped out and the previous state is recovered Features Scale time is 1 s Fit with exponential, based on heuristic model, are used The amplitude of the snapback and the time constant are proportional – can be justified with an heuristic model Corrections Phenomena are fast, they cannot be corrected on the fly but correction curves should be implemented based on measurements USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.29 COMING SOON We now finished the outline of the phenomena involved in the determination of the field homogeneity in a superconducting magnet Geometry of cables Iron effect Deformations Hysteresis Time dependent effect In the next unit we will discuss agreement between the field model and the measurements, agreement between field measurements and expected effects measured on the beam USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.30 REFERENCES General M. N. Wilson, Ch. ? Schmuser, Ch. 6 and 7 L. Bottura, “Field dynamics in superconducting magnets for particle accelerators”, CERN 98-05 (1998). Classes given by A. Jain at USPAS 2006, Unit 7 Ph. D. thesis of A. Verweij Decay and snap-back L. Bottura et al, IEEE Trans. Appl. Supercond. 15 (2005) 1217-20 N. Sammut Ph. D. thesis on the LHC magnets Several works by P. Bauer et al carried out in 2000-2005 at FNAL on Tevatron Complete references on the works on Hera can be found in Schuser book USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.31 ACKNOWLEDGEMENTS L. Bottura and A. Jain for discussions, comments and useful advices USPAS June 2009, Superconducting accelerator magnets Unit 15: Persistent currents and dynamic effects – 15.32